We conducted a sensitivity analysis under four scenarios using data from three Asian studies conducted with low dropout rates to evaluate the efficacy of varenicline in smoking cessation within a scenario where dropouts from the study are dealt with in a manner as unfavourable as possible to varenicline. Pooled odds ratio of continuous abstinence at week 24 was 1.55 (95% confidence interval: 1.03–2.35, random-effect model) under the scenario where dropouts from the varenicline group were dealt with as smokers and those from the placebo group as successful quitters. From the present study, the efficacy of varenicline has been firmly established in the Asian population.